1. NO2 VCD Stratospheric trends: Hemispheric and latitudinal dependence
M. Gil-Ojeda, M. Yela, M. Navarro, C. Robles, F. Hase, B. Funke, O. Puentedura, J. Iglesias, E. Cuevas.

2. Motivation Background concentration of N2O
+ N2O increasing at a steady rate of 2.20%/decade in last 36 years ( )
+ No signs of damping.
+ Lifetime of years
+ Major contribution to NOx in the stratosphere
+ Signature of increase should be observed on long NO2 VCD records

3. INTA Stations for stratospheric monitoring
+ Zenith DOAS at twilight
+ Background stations in clean environments: tropospheric NO2 by pollution not an issue
28ºN. Izaña: FT NO2 (20-40 pptv)
54ºS. Ushuaia -> GAW in the Beagle channel
64ºS. Marambio -> Seymour Island. Antarctic Peninsula
78ºS. Belgrano -> Antarctic mainland
+ NDACC or NDACC applied instruments.

4. Datasets Izaña (28ºN) Ushuaia (55ºS) Marambio (64ºS) Belgrano (28ºN)
Izaña (28ºN)
Ushuaia (55ºS)
Marambio (64ºS)
Belgrano (28ºN)

5. Statistic model (Stolarski,1991, Brunner,2006, Bodeker,1998, 2001
Multiple regression
Statistic model (Stolarski,1991, Brunner,2006, Bodeker,1998, 2001
Other predictor terms
trend
seasonality
Y(t) = NO2 measured column in t time
a = intercept (offset)
Xj = explanatory function (Predictor)
j = Coefficient of the explanatory function Xj
t = time from measurements start (in months)
 = noise in time t
residual
Monthly means data used for the analysis
Consideration of model errors and auto-correlations as described in Stiller et al., 2012.

6. Predictors: Offset (a) Trend (0.t)
Seasonal terms (annual, semiannual, quartely)
Solar activity (10.7cm flux) -> Penticton radio flux data stored in GCMD/NASA
QBO (10hPa and 30 hPa zonal winds over a mean of 3 tropical stations, 2 quasi- orthogonal predictors) -> Berlin_FU.
Stratospheric aerosols optical thickness (Sato et al. 1998) -> Stored in GISS/NASA.
ENSO (multivariate ENSO index MEI) -> (MEI variables: SAT, SST, Wind, Pres, Clouds) -> NOAA (Wolter and Timlin, 1993, 1998) stored in GCMD/NASA
NAO (Monthly normalized index from CPC.NOAA (NCEP)

7. Izaña (28ºN) DOAS pm
Black text: significant to 95%

8. Izaña mean trend (1993-2014) = 7.95%/decade
Izaña data sets
Residuals with trends
trend = ± %/decade
trend = ± 0.65 %/decade
Izaña mean trend ( ) = 7.95%/decade

9. FTIR at Izaña
DOAS over 3 times larger trends than FTIR. ¿Why?
DOAS (am+pm)/2
FTIR
SCIA
OMI

10. FTIR-DOAS-Satellite
All instruments have their largest sensitivity in the stratosphere, but there are differences in vertical sensitivity between instruments
Most of the trends occur in the lower stratosphere
Probably dynamic induced
At 22 km DOAS sensitivity is over 50% larger than FTIR whereas at 34 km the opposite is true

11. MIPAS
NOy Climatology (Funke et al., 2015)
Monthly zonal means (10 deg latitude bins) of NO2+NO+HNO3+2N2O5+ClONO2+HNO4
Full global coverage (independent on solar illumination)

12. NOy Global Trend (B. Funke et al, 2015)
MIPAS
For , both MIPAS and WACCM show hemispherical asymmetric behavior in the lower and middle stratosphere.
WACCM shows a global increase (2-5% per decade) in the polar regions and in the UT, but decrease (2%) in the LM at 30N/30S
WACCM
WACCM
Color shading: significance (95% confidence level) 12

13. All data series available

14. Period
+ Qualitative agreement between DOAS-NO2 and MIPAS NOy
+ Larger trends in GB instruments

15. Possible explanation
+ Chemical -> Total NOy remains without significant trends (Funke, personal communication)
+ Dynamical -> The fact that the trends occur in the low stratosphere together with previous statement provides an evidence of dynamic changes
Change in the strength of the Brewer-Dobson circulation? -> Models agree in an increase in speed in the circulation related to climate change. If this would be the case a decrease in NO2 should be seen on both hemispheres
Displacement of the south subtropical barrier toward higher latitudes after year 2002 (Eckert et al., ACP 2014).

16. Summary
+ A multiple linear regression has been applied to 4 ground-based zenith DOAS data sets of over 20 years (NDACC and NDACC applied unpolluted stations.
+ The Northern Subtropical station (Izaña, 28ºN) shows a positive trend of about 8%/decade, over 3 times larger than the N2O increase (period ).
+ Other instruments (FTIR, and satellite) with higher sensitivity in the upper stratosphere show much less trend, suggesting that changes are mostly restricted to the lower stratosphere where transport dominate to the photochemistry.
+ Spectrometers in Southern Hemisphere show a negative trends, larger than in the NH, and the magnitude of the trend is dependent on the latitude (- 9.8% Ushuaia-55ºS, -15.0% Marambio 64ºS and -19.8% Belgrano-78ºS).
+ MIPAS NOy trend analysis shows qualitatively the same behaviour, although the latitudinal increase is not apparent.
+ Trends are very sensitive to the latitude of the station but also to the selected period -> Trends are not constant in time.

17. Thank you for your attention